Objectives: To analyze the relationship between serum total cholesterol (TC) and all-cause mortality, taking into account various potential confounders.
Design: Population-based prospective cohort study.
Setting: Older Italians residing in the general community.
Participants: Four thousand five hundred twenty-one men and women aged 65-84.
Measurements: Vital status data were available for 1992-95. The hazard ratios of dying for subjects in the second, third, and fourth quartiles compared with the first quartile of TC were computed using Cox proportional hazards, adjusting for lifestyle factors, anthropomorphic and biochemical measures, preexisting medical conditions, and frailty indicators.
Results: Blood samples were obtained from 3,295 (73%) of the participants, of whom 399 died during almost 3 years of follow-up. Low TC was associated with a higher risk of death. Those with TC in the second, third, and fourth quartiles (TC>189 mg/dL or 4.90 mmol/L) had lower hazard ratios (HRs) of death than subjects in the first quartile (0.57, 95% confidence interval (CI) = 0.38-0.87; 0.56, 95% CI = 0.36-0.88; and 0.53, 95% CI = 0.33-0.84, respectively). Few subjects taking lipid-lowering drugs (LLDs) were in the lowest quartile of cholesterol, suggesting that these individuals have low TC values for reasons other than LLD use.
Conclusion: Subjects with low TC levels (<189 mg/dL) are at higher risk of dying even when many related factors have been taken into account. Although more data are needed to clarify the association between TC and all-cause mortality in older individuals, physicians may want to regard very low levels of cholesterol as potential warning signs of occult disease or as signals of rapidly declining health.